YouTube Ad Placement Optimization

This project is designed to optimize YouTube ad placements by leveraging data mining and NLP techniques. The application analyzes video metadata and engagement metrics to provide recommendations for aligning ads with relevant video content.

You can run this project using a Flask web application or the included Jupyter Notebook (SI_final.ipynb).

---

Set Up

1. Prepare Directories

Ensure the following directories are present:

• app.py needs to be in the main directory where the terminal runs

• templates/
  (for HTML templates: index.html and results.html)

• static/plots/
  (for saving plots generated during processing)

2. Set of commands to run in terminal

pip install scikit-learn nltk flask wordcloud gensim
set FLASK_APP=app 
set FLASK_ENV=development
flask run

Open your browser and go to http://127.0.0.1:5000/

3. Using the Application

• Enter your search query and ad keywords on the home page.
• View the top 5 recommended videos, relevance scores, visualizations and download the results as a CSV.

Use jupyter notebook alternatively

• SI final.ipynb This has the sequential processing of code to generate the outputs and also visualize.

Notes

CSV Output:

• The application generates a file named recommended_videos.csv in the project folder containing the recommendations.

Static Plots:

• Generated plots (relevance score bar chart and keyword clouds) are saved in the static/plots/ directory.